Unit Economics for Product Leaders, Part 1: What the Textbooks Get Wrong

Most online resources on unit economics contain errors. Not subtle interpretive disagreements — actual mathematical mistakes that produce wrong answers. After building unit economics models for B2B SaaS products, I want to share what I learned, and what I wish someone had written clearly from the start.

This is Part 1 of a three-part series. This post covers the core formulas, the mistakes that show up everywhere, and the precise definitions that matter. Part 2 connects unit economics to the P&L. Part 3 covers decision frameworks built on top of these metrics.

ARPU Is a Rate, Not a Number

Here's a formula you'll find on dozens of PM blogs and even in university course notes:

LTV = ARPU × Customer Lifetime

Simple enough. But here's where most resources go wrong: they'll define ARPU as " $30 per month" and Customer Lifetime as "10 years," then multiply to get$ 300.

The actual answer is $3,600.

The issue is dimensional. ARPU in these formulas is a rate — dollars per unit of time. Customer Lifetime is a duration. For the units to resolve to dollars, the time basis must match:

If CL is in years, ARPU must be $/year
If CL is in months, ARPU must be $/month
( $/time) × (time) =$ . Always.

The same constraint applies to the equivalent formula:

LTV = ARPU / Churn Rate

Churn rate is a dimensionless ratio per time period (e.g., 10% per year). So ARPU must be denominated in that same period. Monthly ARPU divided by annual churn gives you nonsense: ( $/month) / (1/year) =$ × year/month, which isn't dollars.

This isn't pedantic. I found a widely-shared article that defined ARPU as $30/month, computed Customer Lifetime as 10 years from a 90% annual retention rate, and then wrote `LTV =$ 30 × 0.70 × 10 = $210`. The correct answer, annualizing ARPU first, is `$ 360 × 0.70 × 10 = $2,520`. The error was a factor of 12×.

The fix is simple: before computing anything, write out the units explicitly. If your churn rate and customer lifetime are annual, annualize your ARPU. If they're monthly, use monthly ARPU. Check that the units cancel to produce a dollar figure. If they don't, something is wrong.

The Core Formulas, Done Right

Here are the unit economics formulas that matter for B2B SaaS, with units shown explicitly.

Customer Lifetime

$CL = \frac{1}{Churn\ Rate}$

If annual churn is 20%, CL = 1 / 0.20 = 5 years. If monthly churn is 5%, CL = 1 / 0.05 = 20 months. The output has the same time unit as the input's denominator.

Lifetime Value

Two equivalent formulations:

$LTV = ARPU \times CL$

$LTV = \frac{ARPU}{Churn\ Rate}$

These produce the same result because CL = 1 / Churn Rate. ARPU must share the same time basis as CL or Churn Rate.

Example: ARPU is $120,000/year, annual churn is 25%.

CL = 1 / 0.25 = 4 years
LTV = $120,000 × 4 =$ 480,000
Check: $120,000 / 0.25 =$ 480,000 ✓

Customer Acquisition Cost

$CAC = \frac{Total\ S\&M\ Expense}{Number\ of\ New\ Customers\ Acquired}$

Both numerator and denominator should cover the same period. If you spent $500K on sales and marketing last quarter and acquired 10 customers, CAC =$ 50,000.

LTV/CAC Ratio

$\frac{LTV}{CAC}$

This is a dimensionless ratio ( $/$ = unitless). The benchmark is 3:1 or higher, meaning customer LTV is at least 3× the cost to acquire them. If it's below 3, you're either spending too much to acquire customers, not retaining them long enough, or not charging enough.

Contribution Margin Per Customer

$CM = ARPU - Total\ Variable\ Cost\ Per\ Customer$

Both terms must share the same time period. If ARPU is annual, variable cost must be annual. More on what counts as "variable cost" below.

CAC Payback Period

$CAC\ Payback = \frac{CAC}{Contribution\ Margin\ Per\ Customer\ Per\ Period}$

This is the one formula where the time period of the denominator determines the units of the output. If CM is monthly, payback is in months. If CM is annual, payback is in years. Benchmark: under 12 months for SaaS; 5–7 months is strong.

Critical rule: CAC is in the numerator only. The denominator is the ongoing contribution margin — the recurring surplus that gradually recoups the acquisition investment. Never include CAC in the variable costs used to compute CM. That would double-count it: once as the lump sum you're trying to recover, and again as a drag on the margin that's doing the recovering.

Revenue-Based vs. Margin-Based LTV

The LTV formula above is revenue-based. It tells you the total revenue a customer generates over their lifetime. But revenue isn't profit. If it costs you $40K/year in variable costs to serve a customer who pays$ 120K/year, the revenue-based LTV of $480,000 overstates the economic value of that customer.

A more useful metric for investment decisions:

$LTV_{CM} = Contribution\ Margin\ Per\ Customer\ Per\ Year \times CL$

Using the same example: CM = $120K -$ 40K = $80K/year. LTV(CM) =$ 80K × 4 = $320,000.

This changes the LTV/CAC ratio too. If CAC is $50K:

Revenue-based: $480K /$ 50K = 9.6×
CM-based: $320K /$ 50K = 6.4×

Both are healthy, but the CM-based ratio is the one that reflects actual profit contribution. When evaluating whether a customer segment is worth pursuing, this is the number that matters.

Definitions That Get Conflated

Three revenue metrics that sound similar, mean different things, and get swapped constantly in practice.

ACV (Annual Contract Value): The annualized contract value for a specific client. If a customer signs a 3-year deal worth $900K, their ACV is$ 300K. ACV describes a contract.

ARPU (Average Revenue Per User/Account): The average annual revenue per client across the entire customer base. If you have 50 customers generating $15M in total ARR, ARPU is$ 300K. ARPU describes a statistical average.

ARR (Annual Recurring Revenue): The portfolio-level aggregate. ARPU × number of customers, or the sum of all ACVs. ARR describes the state of the business at a point in time.

In practice, people routinely say "the ARR on this account is $300K" when they mean ACV. It's sloppy but ubiquitous. The distinction that actually matters: ACV is a contract-level fact about one customer. ARR is a portfolio-level snapshot that changes as customers churn, expand, or contract.

One more that's important for forecasting:

NRR (Net Revenue Retention): The percentage of beginning-of-period recurring revenue retained from the same cohort of customers at the end of the period, inclusive of expansion, contraction, and churn.

$NRR = \frac{Beginning\ ARR + Expansion - Contraction - Churned\ ARR}{Beginning\ ARR}$

NRR above 100% means the existing customer base grows revenue even before adding new logos. Below 100% means new sales have to backfill erosion before you can grow. For B2B SaaS with expansion motions (upsell, cross-sell, seat growth), NRR is arguably more important than gross churn for forecasting LTV, because revenue-based LTV implicitly assumes flat ARPU over the customer lifetime. If NRR is 120%, the actual economic value of a customer is significantly higher than what the simple LTV formula produces.

The Weighted LTV Problem

If your pricing has multiple tiers with different ACLs (average customer lifetimes) and different price points, a single blended LTV can be misleading. The right approach is to compute LTV per segment and then weight by the customer mix.

Plan	Price	ACL	Customers
Basic	$600/yr	2 yr	100
Pro	$1,200/yr	3 yr	250
Enterprise	$6,000/yr	5 yr	75

Weighted LTV = [( $600 × 2 × 100) + ($ 1,200 × 3 × 250) + ( $6,000 × 5 × 75)] / (100 + 250 + 75) = **$ 7,765**

If you computed a single blended ARPU ( $4.13M total ARR / 425 customers =$ 9,718) and multiplied by a blended ACL, you'd get a different number because ACL and price aren't independent — enterprise customers tend to stay longer and pay more. The weighted approach preserves this correlation.

Quick Checks Before You Ship a Model

Before presenting any unit economics model — to a stakeholder, an investor, or yourself — run these sanity checks:

Do the units cancel? Write out $/year,$ /month, years, months explicitly. If the result isn't in dollars (for LTV) or months/years (for payback), something is mismatched.
Is ARPU a rate? ARPU should be expressed as $/period, not a cumulative or one-time figure. If your pricing includes one-time setup fees, separate them from the recurring stream.
Is CAC only in the numerator? If you've included acquisition costs in your variable cost calculation, you're double-counting.
Are your time bases consistent? Monthly ARPU with annual churn, or annual ARPU with monthly payback, produces errors.
Are you using revenue-based or CM-based LTV? Know which one you're presenting and why. Revenue-based LTV is fine for top-line narrative. CM-based LTV is what you need for investment decisions.

In Part 2, we'll connect these per-customer metrics to the P&L waterfall — how unit economics and portfolio-level profitability relate, where they diverge, and why the bridge between them is the most important thing most PMs can't explain.